Re: perl code snippet to process Amanda's disklist

Nothing came up in reply to this. So, I wrote a bit of a utility program that does the job for me. 
Call it a hack. It's old style code. It doesn't do everything that others might want and does not 
tie in with Amanda code libraries. But, it works for my case. It may or may not be useful to anyone 
else. I don't really want to formalize it by putting it in the wiki or git or anywhere else. If 
someone does find it useful and wants to adapt it, make changes in it, or put it up somewhere, you 
are welcome to. Note that changes in the format of disklist will break this.

Of course, the "proper" way of doing this would have been to tie it in to the Amanda code library in 
such a way that it would cycle through DLEs by calling Amanda and would acquire any properties of 
each DLE that came from either the disklist or the dump types in amanda.conf. For me, that would 
have been an order of magnitude more work. I would have had to dig into the Amanda code in a fair 
bit of detail to figure out how to do that.

I'm not sure how the Amanda lists are configured with regard to accepting attachments. If the 
attached file does not come through (it is only 3.3KB as a .gz), perhaps someone can advise me as to 
the best way to make it available.

Just for general info, I just added another 23TB Raid6 array to one of my servers. That one is now 
up to about 75TB of storage capacity. The other department is at around 25TB with requests under 
discussion to add more. Labs with the latest digital microscopes can generate 4TB in a week of 
operation. It boggles my mind how we are going to deal with this.


On 3/9/16 5:07 PM, Chris Hoogendyk wrote (to Amanda Users):
> Does anyone have a perl code snippet to process Amanda's disklist?
>
> I found http://wiki.zmanda.com/pod/3.1/Amanda/Disklist.html, but it explicitly says that 
> additional parameters (such as include and exclude) are not reproduced by this package.
>
> Why I am interested in this -- I'm managing four separate 20TB arrays on two different Amanda 
> servers with LTO6 tape libraries and SSD holding "disks" (server grade). I have to break up the 
> arrays into manageable sized DLEs using include and exclude entries that I craft by hand after 
> combing over the results from `du`. Then, a user comes along and shuffles hundreds of giggabytes 
> or even a terrabyte of data from one place to another, and, suddenly, I have a single DLE that is 
> too big and ends up trying to streamtar and gzthrough the pipe directly to tape, which doesn't 
> work all that great.
>
> What I would like to do is have a script on cron that would check the size of each DLE and send me 
> an email alert if any one of them goes above some threshold. Then I can fix it before Amanda runs.
>
> Hey, maybe Amanda already has that capability.   I'm just unaware of it.
>
> I'm running Amanda 3.3.6 on Ubuntu 14.04 LTS.
>
>
> Thank you,

-- 
---------------

Chris Hoogendyk

-
    O__  ---- Systems Administrator
   c/ /'_ --- Biology & Geology Departments
  (*) \(*) -- 317 Morrill Science Center
~~~~~~~~~~ - University of Massachusetts, Amherst

<[email protected]>

---------------

Erdös 4